# Sessions, Liquidity and Trading Calendars Align a trading plan with market hours and information timing. Use this workbook alongside the course. Write your answers before opening the solutions. Practical work is self-reviewed; scored knowledge checks are in the Academy. ## 1. Global sessions ### Trading venue hours Trading hours belong to a venue and instrument. Gold-related markets operate on different schedules; a general statement that gold trades nearly around the clock is not an order-entry schedule for every product. ### Overlapping participation Overlapping participation can change liquidity and information flow, but session labels are broad descriptions rather than guarantees. Observe the actual spread, activity and available data instead of assuming every overlap behaves identically. ### Daylight-saving differences Daylight-saving transitions occur on different dates in different regions. A session associated with a local clock can shift relative to UTC or the learner's clock. Store the named timezone rather than a fixed offset when constructing recurring schedules. ### Local time versus UTC UTC provides a common reference for comparing timestamps. Keep the original timezone information as well when it explains a release or venue rule. A timezone conversion changes the displayed clock, not the underlying event instant. ### Worked example A recurring event is defined in a region that changes its UTC offset seasonally. Hardcoding one UTC hour all year can place the event an hour late during part of the year. ### Independent exercise Build a schedule entry containing the event's named timezone, local time, converted UTC time and the date for which the conversion is valid. My inputs and assumptions: My calculation or decision: Evidence that would change my conclusion: ## 2. Liquidity patterns ### Spread changes Spread is an observable execution cost that can vary across a session. Compare the relevant quote side and size. A historically narrow spread is a reference, not a promise of the next fill. ### Quiet periods and thin books Quiet periods may have fewer participants or less displayed quantity. Thin liquidity can increase the impact of an order or a sudden headline. Low recent volatility does not guarantee low execution risk. ### Open and close behavior Open and close behavior depends on venue mechanisms, order imbalances and information arriving since the preceding session. A pattern around one venue's open may not transfer to another feed's arbitrary daily boundary. ### Why averages do not guarantee today's conditions An average hides variation. A session's typical spread or movement should be accompanied by a range or distribution and a sample definition. Planning exclusively around the mean can ignore the conditions that cause the largest execution problems. ### Worked example A sample has nine observations of spread 1 and one of spread 11. The mean is 2, but a plan assuming every observation costs 2 misses both the common condition and the outlier. ### Independent exercise Calculate the median and mean of that sample and explain why neither alone specifies the cost of the next order. My inputs and assumptions: My calculation or decision: Evidence that would change my conclusion: ## 3. Calendar mechanics ### Economic release time An economic calendar should identify release time, source, jurisdiction and the observation period. Consensus, previous and actual values are different fields. Revisions can change the interpretation of the apparent surprise. ### Holiday schedules Holiday schedules may alter trading, settlement or publication independently. An open market does not imply that every related bank or data source follows its normal schedule. Verify the specific mechanism relevant to the planned activity. ### Contract expiry dates Contract expiry and notice-related dates belong to the actual instrument. A continuous chart can move to a new contract while an account still holds the old one. Record operational dates separately from chart presentation. ### Provider maintenance windows Provider maintenance can interrupt order entry or data delivery. Determine whether existing orders remain active and what official recovery procedure applies. A planned outage should be included in the session plan before positions are opened. ### Worked example A market is open while a related settlement calendar has a holiday. Assuming all dates advance normally can create an incorrect cash-flow expectation even though trading itself continues. ### Independent exercise Create separate calendar fields for trading closure, settlement holiday, economic release and provider maintenance. Explain why one generic holiday flag is insufficient. My inputs and assumptions: My calculation or decision: Evidence that would change my conclusion: ## 4. Session preparation ### Verify current schedules Verify current schedules from the venue, release authority or provider before relying on a saved calendar. A saved template is a starting point, not a current source. Record when the check was performed. ### Mark event-sensitive periods Mark periods when the plan's assumptions may be less reliable, such as a scheduled announcement or thin participation. The response can be reduced activity or no trade; the curriculum does not prescribe an automatic event trade. ### Plan disconnection contingencies A disconnection contingency identifies how state will be checked and which official channel is available. It should not assume that losing a local connection cancels server-side exposure. Rehearse the sequence on a demo account. ### Review missed or delayed information Review missed or delayed information as a process issue. Determine whether the source, conversion, alert or attention failed. Correct the mechanism rather than simply promising to watch more carefully next time. ### Worked example A calendar alert arrives an hour late because a fixed offset was used. The useful correction is a date-aware timezone rule plus a check around transitions, not moving every event manually without preserving its source timezone. ### Independent exercise Write an incident note containing the expected alert, actual alert, cause and a test for the corrected scheduling rule. My inputs and assumptions: My calculation or decision: Evidence that would change my conclusion: ## Course project Build a timezone-correct weekly schedule with holiday and event checks. ### Self-review rubric - Concepts and reasoning: 25% - Calculations, data and evidence: 30% - Process and risk controls: 25% - Limitations and communication: 20% Record one correction and one next practice task. This rubric is not automatically graded. ## Worked solutions ### Exercise 1 The entry should allow a date-aware conversion and retain the original local rule. Recheck around daylight-saving transitions; a fixed numerical offset alone does not describe a timezone's full calendar behavior. ### Exercise 2 The median is 1 and mean is 2. They summarize the sample differently; neither predicts a guaranteed next observation. The outlier and the conditions producing it remain relevant to planning. ### Exercise 3 The four events affect different processes. A single flag cannot explain whether orders can execute, funds settle, information is released or the platform is accessible. Keep the categories separate and cite the applicable schedule. ### Exercise 4 The test should use dates on both sides of the relevant transition and compare against the official local schedule. It should verify the mechanism that failed, not merely confirm one manually corrected event. ## Further reading - https://www.cmegroup.com/education - https://www.lbma.org.uk/prices-and-data/lbma-gold-price/lbma-gold-price